Tourism Management 85 (2021) 104293

Contents lists available at ScienceDirect

Tourism Management
journal homepage: http://www.elsevier.com/locate/tourman

Image, satisfaction, destination and product post-visit behaviours: How do

they relate in emerging destinations?
Catarina Marques a, b, *, Rui Vinhas da Silva a, b, Stanislava Antova c
a
Instituto Universitário de Lisboa (ISCTE-IUL), Business Research Unit (BRU-IUL), Lisboa, Portugal
b
Av. Das Forças Armadas, 1649-026, Lisboa, Portugal
c
Liftoff Mobile, Inc., Moray House, 23 - 31 Great Titchfield St, 2nd floor, Fitzrovia, London W1W 7PA, United Kingdom

A R T I C L E I N F O A B S T R A C T

Keywords: This study proposes a conceptual model that sheds light on how the destination image of emerging tourist
Destination image destinations relates to tourism satisfaction and intention to subsequently recommend the place and purchase its
Unique image products. Destination image is studied through three components – cognitive, affective, and unique. Unique
Emergent destinations
image has been overlooked in previous research and few studies focus on its measurement. This study uses a new
Sofia
Post-visit intentions towards country’s
method of measuring it through text-mining of user-generated blog posts. Findings from a sample of 314 foreign
products visitors to Sofia, Bulgaria, reveal that the affective image influences tourist satisfaction and post-visit behaviour;
the cognitive component has a significant effect on all the constructs, except for tourist satisfaction, whereas
unique image only influences the intention to recommend and purchase destination country products. Joining
together two streams of research, this study also argues that the intention to recommend a destination spot
influences the intention to buy its products.

1. Introduction destinations which are emerging as new frontiers of tourism but are yet
to be developed as tourism products. Still in their infancy as tourism
Destinations, just like products, compete against one another as products, with scarce information provided by tourism agents (as per the
consumers often pick their next holiday destination by opting between induced level of image formation proposed by Gunn (1972)), tourists
places with similar attributes, including climate, sports facilities, safety, perceive the image of an emerging destination differently from that of a
etc. (Qu, Kim, & Im, 2011; Chiadmi, Ladhari, & Souiden, 2017). There mature destination. Consequently, the way tourists evaluate a destina­
are several factors which influence choice of destination, nevertheless tion, their post-visit behaviour and the inter-relationship between all
destination image is one of the most significant decision-making attri­ these variables may differ as well. The current paper focusses on the
butes as it helps to differentiate a place in consumers’ minds (Marchiori study of destination image in the context of emerging destinations, those
& Onder, 2017). Previous research has also evidenced that destination which have not yet been able to successfully create an identity or market
image is related to tourism satisfaction and/or post-visit intentions to strategy, that allows for a unique tourist destination positioning in what
revisit or/and recommend (e.g. Baloglu, Henthorne, & Sahin, 2014; is a cut-throat global tourism market.
Bigné, Sanchez, & Sanchez, 2001; Chen & Tsai, 2007; Chi & Qu, 2008; In the literature, destination image is usually scrutinized as a multi-
Chiu, Zeng, & Cheng, 2016; De Nisco, Mainolfi, Marino, & Napolitano, dimensional construct of two components – cognitive and affective.
2015; Kozak & Rimmington, 2000; Prayag, Hosany, Muskat, & Del Previous research has often overlooked a third important component –
Chiappa, 2017; Qu et al., 2011), making it a good tool to predict tourist the unique destination image. The latter has proved to have a stronger
behaviour, whilst creating meaningful and tailored offerings. influence on forming the overall image of a destination than the affec­
The destinations which normally grab attention in academic research tive cue and that is important as a key differentiator of the destination
are mostly places where over-tourism is already setting in, and mar­ proposition (Qu et al., 2011). There are only a small number of studies
keting strategies are already in place (e.g. studies of Chen and Tsai which examine the perceptions of the unique features of a destination (e.
(2007), Chi and Qu (2008), Elliot, Papadopoulos, and Kim (2011), g. Qu et al., 2011; Echtner & Ritchie, 1993), and even fewer which study
Baloglu et al. (2014) and Prayag et al. (2017)). However, there are also their measurement. Identifying the unique elements of a destination

* Corresponding author. Instituto Universitário de Lisboa (ISCTE-IUL), Business Research Unit (BRU-IUL), Lisboa, Portugal.
E-mail addresses: [email protected] (C. Marques), [email protected] (R. Vinhas da Silva), [email protected] (S. Antova).

https://doi.org/10.1016/j.tourman.2021.104293
Received 30 March 2019; Received in revised form 31 May 2020; Accepted 29 January 2021
Available online 10 February 2021
0261-5177/© 2021 Elsevier Ltd. All rights reserved.
C. Marques et al. Tourism Management 85 (2021) 104293

should be the starting point of every positioning strategy as these ele­ post-visit behaviours. Therefore, the affective, cognitive, and unique
ments are critical for the establishment of meaningful differentiation image are examined as components of the overall destination image
(Crompton, Fakeye, & Lue, 1992); this is of the utmost for destinations in using data from international tourists visiting Sofia.
the early stages of the life cycle (Butler, 1980). Unique image has often been overlooked in previous studies. In
Previous research has shown that when travelling abroad, in­ addition, whereas the measurement of the cognitive and affective im­
dividuals tend to augment the information search for local products and ages is well established in the literature, this is not true for the unique
modify their attitudes towards them (Hallberg, 2005). Having become image as the uniqueness of one destination is completely different from
acquainted with a destination country’s products and having grown all of the others. The current study augments the literature on unique
fond of them, upon returning home tourists are predisposed to buy these destination image by studying the perceptions held of the unique fea­
products as well as telling others about them; this may be construed as a tures of Sofia by using a method to extract measures based on text
desirable travel outcome as it impacts favourably on the exports of the mining of user-generated content (UGC).
visited country (Kotler & Gertner, 2004). Although the extensive body of
literature within the tourism destination image and the 2. Literature review and hypotheses development
country-of-origin image areas, both providing support for the notion
that the image of a place has significant influence on consumer behav­ 2.1. Destination image
iour, research empirically testing combined elements from the two fields
is scarce, in particular, regarding the influence of destination image on Destination image has been widely defined as impressions or per­
tourist intentions and consumption-related behaviour concerning ceptions of a place. Hunt (1975) proposed that image is a perception
products from these countries/destinations (Elliot et al., 2011; De Nisco held by potential tourists about a specific destination whereas Um and
et al., 2015; De Nisco, Elliot, & Papadopoulos, 2016). The current study Crompton (1990) described it as a holistic construct. It is also often
intends, not only to incorporate the intention to recommend and to buy referred to as the mental picture which an individual has of a certain
products made in the country of destination in the model, as previous place (Bigné et al., 2001; Kotler, Haider, & Rein, 1993). Tasci, Gartner,
studies do, but also relate it with the destination image components and and Cavusgil (2007) define it as “an interactive system of thoughts,
intentions to recommend the tourism destination, relationships not yet opinions, feelings, visualizations, and intentions towards a destination”
studied. Studying these relationships is critical to the understanding of (Tasci et al., 2007, p. 200), stressing on the complex nature of image and
the indirect impacts of tourism on countries’ economic development, its components.
and more specifically to market development for export products. Most researchers have conceptualized destination image as a multi­
In summary, the current study contributes to a better understanding dimensional construct of two components: cognitive and affective (e.g.
of destination image and its impacts on tourist satisfaction and post-visit Baloglu & McCleary, 1999; Hosany, Ekinci, & Uysal, 2007; Baloglu
intentions. A conceptual model is proposed and analysed in the context et al., 2014; Lopes, 2011). The cognitive component is related to the
of emerging destinations in order to provide answers to the following beliefs and knowledge which one holds about the attributes of a certain
research questions: place, that is, cognitive evaluation of image is based on factual knowl­
edge, personal beliefs, meanings, and memories. On the other hand, the
(1) How does the image of an emerging destination influence affective component is related to the emotions and feelings, which a
tourist satisfaction, intention to recommend the destination, person holds about a place (Baloglu et al., 2014), and it can be either
and post-visit intentions towards country’s products? favourable, unfavourable, or neutral (Arslanova, Agapito, & Pinto,
2017). Affect is expressed with positive or negative feelings with varying
More precisely, this study is aimed at explaining how destination intensity. On the one end of the spectrum are emotions, such as love and
image, tourist satisfaction, and the post-visit intentions of foreign visi­ anger, followed by feelings such as satisfaction and frustrations, and
tors towards the destination itself and the country’s products are related moods like boredom or relaxation. At the other end are evaluations such
in a context of emerging destinations. A conceptual model is proposed as liking and disliking (Peter & Olson, 1999; Tasci et al., 2007).
which adds some variables and relationships to already tested According to Kim, Mckercher, and Lee (2009), the affective
frameworks. component is more volatile than the cognitive one as it is based on
Sofia, Bulgaria, is chosen as the emerging tourism destination. Sofia emotional situations. The authors also found out that cognitive image
is one of the emerging cities where the number of foreign visitors is tends to last longer as it is subject to previously formed knowledge.
constantly growing, although a defined vision and strategy for tourism Baloglu (1999) and Baloglu et al. (2014) found out that the influence of
have not yet been implemented. The city was ranked by Mastercard’s these components tends to differ between visitors and non-visitors. The
Global Destination Cities Index study (Mastercard, 2017) as the third authors suggest that the cognitive evaluation of an image is a more
fastest growing European city of destination in terms of international dominant factor among non-visitors, and the affective one becomes
overnight visitors from 2009 to 2016. This study also discusses practical stronger once a tourist visits the place.
implications for emerging destination marketeers, in particular Sofia In their three-dimensional framework, Echtner and Ritchie (1991)
marketeers and tourism managers in order to successfully create an suggested that the destination image can have common characteristics
identity or market strategy to position the city as a unique tourist as well as unique features i.e. features which sets a place apart from all
destination in the global and increasingly competitive tourism market. the others (e.g. local gastronomy, experiences, history, traditions, etc.).
All these unique characteristics grouped together form the unique
(2) Does unique image play a key role as a component of desti­ image, a set of both cognitive and affective characteristics which are
nation image? unique and typical for a certain place and which help to differentiate it
from other destinations in tourists’ minds (Qu et al., 2011). Contrary to
Moreso than mature destinations, emerging ones need that their the cognitive and the affective components, the unique component is not
unique characteristics and products be positioned and promoted in such that broadly studied in the destination image literature. An example of a
a way as to attract tourists (Qu et al., 2011). In the early stages of study which included unique image is the one by Qu et al. (2011) who
tourism development, tourists do not have a clear image of the place of created a conceptual model of overall destination image consisting of
destination (Butler, 1980). As a result, the knowledge and affection cognitive, affective, and unique images. The authors found out that the
people have for emerging destinations can be quite limited, which may unique attributes of a destination are critical to forming the overall
make the unique features that are found and experienced during the trip, image in consumers’ minds, and that they are more important than the
much more influential factors when assessing both their stay and affective component. Consequently, a strong unique image is more likely

2
C. Marques et al. Tourism Management 85 (2021) 104293

to lead to a more favourable overall image of a destination. H2. Affective destination image and tourist satisfaction are positively
related.

2.2. Tourism satisfaction H3. Unique destination image and tourist satisfaction are positively
related.
There is generally a debate on whether satisfaction is a cognitive
judgment (e.g. Chadee & Mattsson, 1996), an affective state (e.g. 2.3. Post-trip intention to recommend a destination
Westbrook, 1980), or a combination of both (e.g. Oliver, 1993; Del
Bosque & Martín, 2008; Martínez Caro & Martínez García, 2007; Bigné, Intention to recommend has been studied extensively in the mar­
Andreu, & Gnoth, 2005; Chen & Phou, 2013). As customer satisfaction keting and tourism literature. Confente (2015) did a critical review of 46
literature evolved, more and more scholars started to investigate studies on word-of-mouth communications (WOM) conducted between
customer satisfaction not only with the cognitive approach, but also 1987 and 2013. Some of the studies were based on the influence of WOM
from a more affective perspective (Del Bosque & Martín, 2008; Oliver, on travel decisions (e.g. Leach, Liu, & Winsor, 2008; Murphy, Mascardo,
1993; Martínez Caro & Martínez García, 2007; Chen & Phou, 2013). & Benckendorff, 2007), online WOM and the features of online reviews
They started to include different emotional variables into the concep­ (e.g. Stringam & Gerdes, 2010; Racherla, Connolly, & Christodoulidou,
tualization of satisfaction, particularly in the service and tourism 2013; Park & Allen, 2013), the eWOM influence on travel decisions (e.g.
research, since these industries have an experimental nature and the Patterson, 2007), and the motivations of consumers to search for WOM
consumers’ feelings are a big part of their experience (Bigné et al., 2005; (e.g. Kim, Mattila, & Baloglu, 2011).
Martínez Caro & Martínez García, 2007). Current definition of WOM includes not only all types of informal
Oliver (1993), adding to his cognitive paradigm an affective interpersonal communications from people that the consumers know, as
component, suggests that satisfaction is also influenced by negative and in the past definitions, but also sources from online platforms and other
positive emotions. According to this affective approach, satisfaction is influencers which are not related to the brand or the sellers (Confente,
defined as the consumer’s fulfilment response – whether or not the 2015). The emergence of social media networks, blogs, websites,
product/service was able to provide the desired level of pleasure (Oliver, recommendation sites, and virtual communities made it possible for
1993). The latter, together with arousal, form the two dimensions of people to share and read each other’s opinions about different products,
emotions (Bigné et al., 2005), where arousal is the extent to which a services, and experiences, which led to the emergence of the concept of
person feels activated or stimulated (Russell & Pratt, 1980) and pleasure eWOM (electronic word-of-mouth) (Chevalier & Mayzlin, 2006; Con­
is the degree to which a person feels good, joyful, or happy (Bigné et al., fente, 2015). In the tourism industry, websites, such as TripAdvisor,
2005). Zomato, and FourSquare, offer a platform for consumers to share their
In recent years, the mostly adopted approach is a combination of experiences and perceptions about destinations, hotels, restaurants,
both – the cognitive-affective model – which suggests that satisfaction is tours, etc. (Sigala, Christou, & Gretzel, 2012). Positive reviews online
influenced by both the cognitive judgments of the consumers, as well as work as free advertising for destinations and tourism providers, however
their emotional response to the experience (e.g. Bigné et al., 2005; Chen negative comments online can have the opposite effect and damage the
& Phou, 2013; Del Bosque & Martín, 2008; Martínez Caro & Martínez image and reputation of the company/destination (Chen & Law, 2016).
García, 2007; Oliver, 1993). In this light, the mental processes of Although eWOM is a relatively new area of research, it has been gaining
assessing the experience are done by the cognitive system, whereas popularity over the last decade (Chen & Law, 2016). eWOM studies in
emotions are related to the consumer’s feelings towards the service (Del hospitality and tourism management, in the 2008–2014 period, showed
Bosque & Martín, 2008). that research was generally related to three topics: the nature and
Tourism satisfaction has been extensively studied in the literature characteristics of eWOM, antecedents of eWOM, and its impact (Chen &
regarding destination image. The positive relationship between the two Law, 2016).
constructs is well established in past studies (e. g. Bigné et al., 2001; The literature suggests that both destination image and satisfaction
Kozak & Rimmington, 2000; Chen & Tsai, 2007; Aliman, Hashim, are important antecedents of intention to recommend a place (Bigné
Harudin, & Wahid, 2016). One stream of research is dedicated to et al., 2001; De Nisco et al., 2015; Prayag et al., 2017). Intention to
examining the role of the predetermined destination image in the for­ recommend is one of the most important behavioural outcomes trig­
mation of expectations prior to the trip, which are later used to evaluate gered by destination image (e.g. Baloglu et al. 2014; Bigné et al., 2001;
satisfaction by comparing them with the actual experience. Tourists Chen & Tsai, 2007; Qu et al., 2011; Chiu et al., 2016; Prayag et al.,
depend on their previous knowledge about a destination in order to be 2017). Baloglu et al. (2014) showed that cognitive, affective, and overall
able to evaluate whether it can satisfy their needs (Chen & Phou, 2013). destination images are all predictors of intention to recommend for
Another stream of research is dedicated to proving the positive in­ first-time visitors, whereas repeat visitors do not rely on cognitive
fluence of destination image on tourist satisfaction (e.g. Bigné et al., destination image while recommending it.
2001; Chiu et al., 2016; Ramseook-Munhurrun, Seebaluck, & Naidoo, Tourism literature has also demonstrated that tourists with higher
2015; Chi & Qu, 2008; Loureiro & González, 2008; Aliman et al., 2016). levels of satisfaction are more willing to spread positive WOM (e.g. Chen
Bigné et al. (2001) report that destination image directly influences & Tsai, 2007; Chi & Qu, 2008; Chiu et al., 2016; Prayag et al., 2017;
perceived quality and satisfaction. Similarly, Ramseook-Munhurrun Yoon & Uysal, 2005; Bigné et al., 2001). This was also shown in the
et al. (2015) show that destination image is a direct determinant of study of Phillips, Wolfe, Hodur, and Leistritz (2013), which indicates
satisfaction, and Aliman et al. (2016) demonstrate that the higher the that both attribute-based and overall satisfaction are predictors of pos­
destination image which tourists hold, the higher the satisfaction levels. itive WOM. Moreover, Ozturk and Gogtas (2016) researched how the
Chiu et al. (2016) further evidence that both the cognitive and the af­ satisfaction of cruise visitors with a destination can influence their
fective components of destination image influence satisfaction. More­ intention to recommend it to people in their social and professional
over, destination image directly influences attribute-based satisfaction, network. They suggested that satisfaction with a destination has a pos­
and destination image and attribute satisfaction are both direct ante­ itive influence on the word-of-mouth recommendations intentions of the
cedents of overall satisfaction (Chi & Qu, 2008). tourists.
Having the aforementioned research in mind, the following hy­ Based on the aforementioned literature, four hypotheses are formu­
potheses are suggested: lated as follows:
H1. Cognitive destination image and tourist satisfaction are positively H4. Cognitive destination image and intention to recommend are
related. positively related.

3
C. Marques et al. Tourism Management 85 (2021) 104293

H5. Affective destination image and intention to recommend are 2.5. Conceptual model
positively related.
Fig. 1 presents the conceptual model and the twelve hypotheses. The
H6. Unique destination image and intention to recommend are posi­
relationships between the three components of destination image and
tively related.
intention to recommend and post-visit intention towards country’s
H7. Tourism satisfaction and intention to recommend are positively products will be analysed through the mediation of tourist satisfaction.
related. The model will be estimated in the context of emerging destinations,
but it is also valid for mature destinations.
Tourism destinations do change over time, mainly due to changes in
2.4. Post-visit intentions towards products made in the visited country
tourist preferences and needs, in infrastructure and facilities and also in
whatever accounted for the destination’s popularity (Butler, 1980).
In general, there are two major concepts which have been studied in
Attractiveness thus varies for different typologies of tourists as the
the literature with regards to intention towards a country’s products:
destination itself evolves. Several models describing destination devel­
country-of-origin image (COI) and tourism destination image (TDI) (e.g.
opment have been put forth in the literature (Hunt & Stronza, 2014);
Elliot et al., 2011; Lee & Lockshin, 2012, Papadopoulos & Heslop, 1986;
however, Butler’s (1980) Tourism Area Life Cycle (TALC) model has
Hallberg, 2005). Even though they have emerged as separate constructs,
been the most used. The TALC model suggests that there are five stages
both are focused on studying how the image of a specific coun­
which a tourism spot undergoes: “exploration”, “involvement”, “devel­
try/destination could impact on consumer behaviour (Elliot et al.,
opment”, “consolidation”, “stagnation”, and “decline” or
2011). Interactions between the tourism, product, and overall image of a
“rejuvenation”.
place have been studied in a very few studies. For instance, Elliot et al.
An emerging destination would theoretically be assumed to be in the
(2011) first proposed an integrative model, encompassing both the
first two stages, “exploration” and “involvement”. It is a destination
product and tourism sides of a country image. They studied how the
where tourism products have not been fully developed and there may be
cognitive and affective country image and products and destination fa­
little awareness of the destination by the market (Butler, 1980). These
miliarity impact the products and destination beliefs as well as their
specific characteristics account for the need to understand the tourists’
receptivity (i.e. the intentions towards the products and the destination).
behaviour in order to define a broad strategic marketing approach,
Later, De Nisco et al. (2015, 2016) added tourism satisfaction to the
namely for the promotion of emerging tourist destinations.
latter model, in order to understand how it influences the intention to­
wards tourism and products made in the country of travel. In these
3. Methodological approach
studies, the cognitive and affective images are studied as components of
the general country image while destination beliefs/image was consid­
The adopted methodology consists of both a qualitative and a
ered as a whole. However, the way the components of destination image
quantitative study. Firstly, a qualitative text analysis of online blog posts
as well as the destination WOM affect the post-visit intentions towards
was performed to determine the measures of the construct of unique
the destination’s products have not been paid the same attention.
image. The second phase was focused on the quantitative analysis,
Nevertheless, as several studies have shown that attitudinal loyalty
which included the questionnaire design, sample collection and data
constructs, including the desire and intention to revisit the destination,
analysis. Measurement of the other two destination image components,
WOM and recommendation are highly intercorrelated (Konecnik &
cognitive and affective, were developed using measures of previous
Gartner, 2007), we hypothesise that whoever intends to recommend a
research.
destination also intends to buy the country’s products.
Insetting the context for the qualitative text analysis, it is deemed
In terms of COI, previous studies have demonstrated that customers
appropriate to briefly elaborate on the city of Sofia. Thus, a brief
tend to have a more favourable opinion about products made in coun­
description is presented.
tries with positive images (Chattalas, Kramer, & Takada, 2008; Lee &
Lockshin, 2012). Bilkey and Nes (1982) found that the country image
3.1. Destination Sofia, Bulgaria
alone is an influential factor for consumers to evaluate the quality of
products which they have never tried or purchased before. In TDI,
Sofia is the capital of Bulgaria and its largest city, located in the
Papadopoulos and Heslop (1986) studied how the opinion of Canadians
western part of the country. As of 2015, the city had an estimated
about country’s products varies between those who have visited the
population of 1,260,120 people (World Population Review, 2017). It is
country and those who have not. The authors found a significant dif­
the main administrative, industrial and transportation centre of
ference in the opinions of the two groups, proving that visiting a desti­
Bulgaria. Moreover, it is one of the oldest cities in the world, and as such,
nation changes how the country’s products are perceived. This was
it has a rich cultural heritage with 1400 monuments. Sofia is also
further confirmed by Hallberg (2005) who found that travel experiences
endowed with a rich cultural life with several film, music, and art fes­
can cause changes in the travellers’ intentions towards products asso­
tivals being hosted there every year.
ciated with the visited country.
Sofia is surrounded by three mountains – Vitosha to the south, Lyulin
According to the reviewed studies, the following hypotheses are
to the west, and the Balkan Range to the north, offering natural attrac­
defined:
tions and hiking opportunities. Vitosha Natural Park offers opportunities
H8. Cognitive destination image and post-visit intentions towards for several adventure and other extreme sports. The Bulgarian govern­
country’s products are positively related. ment has taken action to building the image of the city as a sports
destination. Thus, several marathons and sports events were hosted in
H9. Affective destination image and post-visit intentions towards
the past few years which have led to Sofia earning the title of European
country’s products are positively related.
Capital of Sport 2018.
H10. Unique destination image and post-visit intentions towards International tourism in Sofia has significantly increased over the
country’s products are positively related. last years. The Bulgarian capital was ranked the third European city, in
terms of growth in incoming international tourists for the period
H11. Tourism satisfaction and post-visit intentions towards country’s
2009–2016. The city was visited by 1.19 million tourists in 2016, which
products are positively related.
is a growth of 18% in comparison with the previous year (Mastercard,
H12. Intention to recommend a tourism destination and post-visit 2017). According to the Bulgarian financial media Capital (Stoilova,
intention towards country’s products are positively related. 2016), one reason for this was the fact that low-cost airline RyanAir

4
C. Marques et al. Tourism Management 85 (2021) 104293

Fig. 1. Research model and hypotheses.

started operating with 21 flights from Sofia in September 2016. Most of Secondly, coding suggested by the automated algorithm was subject to
the tourists spend a weekend in the city (Stoilova, 2016). Tourists human judgment, carried out by a researcher and professionals of the
recognise that the city has a good balance between price and quality of Sofia Tourism. In this second stage, the potential coding was carefully
food, accommodation, and nightlife and they are attracted to it because analysed and manually revised as the concept under analysis was to
of its “exotic unfamiliarity” (Stoilova, 2016). Sofia needs therefore to identify unique features. Some studies (e.g. Sotiriadou, Brouwers, & Le,
work hard to create a destination brand and identity. According to the 2014) have brought it up that the manual data analysis could be sub­
Sofia municipality, the aim is to position Sofia as a “modern European jective, but also more engaging for the researchers. Redundant nodes
city focused on cultural tourism” (Stoilova, 2016). were deleted, and some nodes were reclassified into different categori­
zation (e.g. names of communist buildings and institutions were put into
3.2. Phase I: qualitative text analysis a folder called Communist Heritage) with the collaboration of two
tourism professionals. In summary, the second stage aimed to reorganize
The first phase of the research process was to identify and create a list the coded data and generate patterns and connections. The final coding
of the unique features of Sofia, which were later used to measure the was also validated by the tourism professionals to ensure that the clas­
perception of unique image in the questionnaire. As we did not assume sification procedure was reliable.
the existence of pre-established categories, but that these would emerge Common codes were grouped into 10 different nodes and 24 sub-
from the data, an indutive approach was conducted; Therefore, in order nodes, presented in Table 1, featuring both cognitive and affective fea­
to identify the measurement items, a text mining analysis of UGC was tures which were considered as unique by the blog writers.
performed using NVivo 12. Text mining is “a computer assisted tech­
nique that is equipped with the capability to extract information and
3.3. Phase II: quantitative analysis
trends from large amounts of textual data, giving an overview of the
main issues discussed” (Aureli, 2017, p. 4). Text mining has found wide
3.3.1. Sample and data collection procedure
application in a number of fields, including academic and industry
To test the proposed model, a survey was conducted on a sample of
research, social media and web analysis, business intelligence, etc.
international tourists. The data collection took place over the course of
(Talib, Hanif, Ayesha, & Fatima, 2016). The process of text mining offers
four weeks in the months March and April of 2018. The questionnaire
different techniques and tools to extract information from a text,
including summarization, classification, clustering, natural language
Table 1
processing, etc. The latter could be used for opinion mining, feature
Nodes and sub-nodes extracted from UGC.
extraction, sentiment, predictive, and trend analysis (Talib et al., 2016).
The techniques used include text summarization and classification. Nodes Sub-nodes
This research employed a netnography approach to gather online Attractions
data from UCG in the form of blog posts about Sofia (Kozinets, 2002). Alexander Nevsky Cathedral
The final data set included a total of 72 blog posts published between Ivan Vazov National Theatre
National Palace of Culture
2015 and 2018. All of the posts were extracted from individual blogs Slaveykov Square open book market
written by foreign visitors in English, and the total number of words Square of Tolerance
accounted for 86,395. The type of blogs at hand was on travelling, and Statue of Saint Sofia
the selected posts included guides, tips, and advice about what to do Vitosha Boulevard
Vitosha Mountain
while visiting the city. Moreover, they were based on the personal
Adjectives
opinion and experience of the bloggers, who own the websites. After the Multicultural city
data was collected, it was corrected for spelling mistakes by running a One of the oldest capitals in Europe
spell check and the names of tourist attractions were checked to make Quirky and charming city
sure that they were written in the same manner. City of contrasts
The cheapest capital in Europe
After importing the data in NVivo, text was coded into different Underrated tourism destination
nodes. Coding is an essential part of qualitative analyses as it helps to Communist History and Heritage
gather all of the quotes and references about a particular concept into a Soviet Architecture and Buildings
separate folder for further exploration. This separate folder is called a Delicious and distinctive traditional cuisine
Vegetarian and vegan-friendly destination
node and it serves as a sort of a container for different themes and
Availability of hot mineral water
concepts. The process of coding into nodes helps to identify patterns and Selection of Free Walking Tours
generate ideas in the research materials (Wong, 2008). In this study, the Great Wi-Fi connectivity
coding procedure was conducted both automatically and manually Sofia nightlife
separately in two stages. Firstly, text mining was used to structure the Squat shops
Street art
input text and develop the initial coding scheme from the data.

5
C. Marques et al. Tourism Management 85 (2021) 104293

was administered both face-to-face and online. Face-to-face was done Fig. 1. AMOS 25 was used to estimate the model and the maximum
predominantly by asking foreign visitors to fill it on paper or on a tablet likelihood estimation method was adopted.
in the departure zones of Sofia Airport before leaving the city. Addi­ The two-stage modelling process proposed by Anderson and Gerbing
tionally, a smaller number of questionnaires were distributed at the Visit (1988) was considered. In the first stage, confirmatory factor analysis
Sofia’s tourism office and other institutions, such as museums, galleries, (CFA) was conducted in order to properly evaluate the overall mea­
hostels, etc. for self-administration. The questionnaire was also distrib­ surement model (Anderson and Gerbing, 1988) regarding to the item
uted online by contacting people who left social media reviews about the and construct reliability, and convergent and discriminant validity. In
city’s attractions during the period of the study. The websites used the second stage, the structural component of the model was estimated
include Facebook, TripAdvisor, Twitter, and Instagram. The question­ based on the overall measurement model results. In addition to the es­
naire itself included 18 questions relating to decision-making factors, timates of the hypothesised relationships (direct effects between two
destination image, tourist satisfaction, intention to recommend the city, constructs), indirect and total effects were estimated to assess whether
intentions toward the recommendation and purchase of Bulgarian the tourism satisfaction and intention to recommend factors could be
products, socio-demographics of respondents, and characterization of considered as mediators. The bias-corrected (BC) bootstrap with 1000
the trip. Appendix A presents the questionnaire. samples was adopted to produce more accurate confidence limits for the
A pre-test was performed on a sample of 58 prior to collecting the indirect effects (MacKinnon, Lockwood, & Williams, 2004; Cheung &
final sample. The internal consistency analysis of each construct, more Lau, 2008). When examining mediation effects with latent variables,
specifically the Cronbach’s alpha, was examined in order to see how Cheung and Lau (2008) recommend the use of SEM to control for the
closely related the items are as a group. Alpha increases as the corre­ effects of measurement errors. Therefore, the BC bootstrap confidence
lations among items increase, therefore it is known as a measure of the intervals were used in examining the significance of both indirect and
internal consistency of instrument reliability. The reliability scores of total effects. The Bollen-Stine bootstrap procedure with 2000 samples
Cronbach’s alpha for each construct were between 0.808 and 0.934 and was also used to obtain a goodness-of-fit statistic (Bollen & Stine, 1990)
were considered as good and excellent reliability, respectively. appropriate to non-normal data.
A quota sampling method was applied. This method is used when Along with the Chi-Square test, the Comparative Fit Index (CFI), the
there is no sampling frame available. The strata were defined by country Tucker-Lewis Index (TLI), and the Root Mean Squared Error of
of origin and age groups, in proportion to the population distribution. As Approximation (RMSEA) were used to assess the model-data fit. The
official data about the tourists’ characterization in Sofia was not found, selection of these indices was made based on Hair et al.’s (2014)
unofficial data about the country of origin and age from tourist centres recommendation to have at least one absolute and one incremental fit
was used instead. In order to avoid bias and diversify the data, the indices.
interviewer aimed to collect data on different dates with different flight
destinations, as well as on various places in Sofia. At the end of the 4. Results
survey, a total of 314 respondents from 55 countries constitute the
sample. 4.1. Sample characterization

3.3.2. Measures of the model constructs The sample is composed by 314 respondents, from which 50.6% are
Adapted scales from previous studies were used to create measures male. The age of the respondents is distributed mostly between 18 and
for each one of the model constructs. They were selected based on the 45 years old, with a higher concentration in the “25–34” range (34.7%).
specific characteristics of the destination. The unique image was The majority of the respondents have Bachelor (35.7%) or Master De­
measured using the multi-attribute approach (Echtner & Ritchie, 1991), gree (33.4%), average income (41.7%) or above average income
through the 24 features obtained from the qualitative text analysis of the (36.0%); 25% come from Southern Europe whereas about 20% come
blog posts. These included both cognitive and affective variables which from Western Europe. In terms of separate countries, most tourists came
were considered unique by the blog posts writers. The cognitive image from the United Kingdom (11.2%), Italy (10.2%), United States of
measurement was developed by selecting 29 items from previous studies America (6.7%), Spain (6.3%), Germany (5.7%), and France (5.7%).
(e.g. Baloglu & McCleary, 1999; Basaran, 2016; Beerli & Martín, 2004; The purpose of visiting Sofia was mostly Holiday/leisure (50%),
Qu et al., 2011; Ramseook-Munhurrun et al., 2015; Stylos, Vassiliadis, followed by Business (18.5%), and visiting family and friends (17.5%).
Bellou, & Andronikidis, 2016). One additional item “Availability of The duration of the stay was mostly more than three days (29.9%), three
organized sightseeing tours” was added by the researchers. Similarly, nights (25.8%), or one week (22%). Most of the respondents travelled
the affective image was measured by 14 variables, four items (‘Pleasant’, with their friends (28%), solo (26.4%), or as a couple (23.2%). They
‘Exciting’, ‘Relaxing’ and ‘Arousing’) developed by Baloglu and chose to travel to Sofia mainly because the trip cost (mean = 3.93; sd =
McCleary (1999) and the remaining added by the researchers. 1.19, in a 5-point Likert scale where 1-not important/5-very important),
Tourism satisfaction was measured by five items which were taken getting an opportunity (mean = 3.81; sd = 1.26), or due to friends and
from Yoon and Uysal (2005), Bigné et al. (2001), and De Nisco et al. family counsel (mean = 3.11; sd = 1.45).
(2015). One item (“My visit to Sofia was a wonderful surprise”) was also
added by the researchers. Intention to recommend was measured by six 4.2. Overall measurement model
items, relating to the intention to recommend face-to-face and intention
to recommend online. The five measures for the post-visit intentions CFA were conducted in order to test the overall measurement model.
towards Bulgarian products were adapted from De Nisco et al. (2015). Several items of the measurement scales were dropped from the analysis
These statements as well as those related to the unique and cognitive because of lower item reliability or because they did not assess the
features of Sofia were measured using a 5-point Likert scale ranging model convergent validity criterion. Table 2 presents some measures of
from 1 (strongly disagree) to 5 (strongly agree) whereas the those the overall measurement model. All items and constructs exhibit reli­
related to affective image, tourist satisfaction and intention to recom­ ability, as can be proved by the standardized loadings and CR values,
mend Sofia were measured based on a 7-point Likert scale ranging from respectively. Regarding the convergent validity, the AVE value for all
1 (strongly disagree) to 7 (strongly agree). latent constructs exceeds the minimum required value of 0.50 (Bagozzi
& Yi, 1988), except that of Unique Image. However, as AVE is a more
3.3.3. Statistical analysis conservative measure than CR, Fornell and Larcker (1981) argue that if a
Structural Equation Modelling (SEM) was used to test the hypoth­ construct has the AVE value lower than 0.5, but it is a strongly reliable
esised relationships and to validate the proposed model displayed in construct (i.e. its CR value is higher than 0.6 as is the case of the Unique

6
C. Marques et al. Tourism Management 85 (2021) 104293

Table 2
Descriptive and reliability measures of the overall measurement model items.
Items Loading Mean Std. Deviation Skewness Kurtosis

Cognitive Image (AVE = 0.552; CR = 0.892)

Good quality of accommodations 0.627 3.777 0.887 − 0.622 0.386
Good variety of accommodations 0.581 3.793 0.860 − 0.586 0.540
Good quality of restaurants 0.913 4.172 0.836 − 1.153 1.790
Good variety of restaurants and cuisines 0.771 4.140 0.823 − 0.850 0.693
Appealing local food cuisine 0.785 4.178 0.932 − 1.285 1.659
Variety of products that promote local culture 0.517 3.898 0.888 − 0.595 0.226
Good value for money 0.512 4.431 0.752 − 1.296 1.574
Affective Image (AVE = 0.514; CR = 0.840)
Pleasant 0.875 5.825 1.250 − 1.817 4.178
Exciting 0.775 5.366 1.262 − 0.914 0.880
Relaxing 0.694 5.188 1.368 − 0.761 0.068
Interesting 0.853 6.041 1.159 − 1.906 4.748
Cosy 0.715 5.268 1.354 − 0.749 0.136
Unique Image (AVE = 0.344; CR = 0.807)
Ivan Vazov National Theatre 0.521 3.885 1.033 − 0.693 − 0.031
Square of Tolerance 0.525 3.529 1.012 − 0.254 − 0.329
Slaveykov Square open book market 0.568 3.363 0.973 − 0.299 − 0.020
Multicultural city 0.543 3.436 1.176 − 0.330 − 0.764
City of contrasts 0.515 3.869 1.004 − 0.666 − 0.055
Street art 0.642 3.334 0.998 − 0.205 − 0.434
Squat shops 0.602 3.118 0.999 − 0.218 0.004
Sofia nightlife 0.567 3.459 1.039 − 0.334 − 0.166
Tourist Satisfaction (AVE = 0.522; CR = 0.845)
My visit to Sofia is worth my time and effort. 0.811 6.127 1.216 − 2.070 5.012
My travel experience to Sofia exceeded my expectations 0.874 5.659 1.419 − 1.211 1.148
In comparison with other similar places I’ve visited before, Sofia is a better destination for tourism 0.706 4.557 1.610 − 0.340 − 0.518
My visit to Sofia was a wonderful surprise 0.874 5.344 1.547 − 0.876 0.125
Overall, I am satisfied with my travel experience in Sofia 0.885 6.086 1.234 − 1.931 4.191
Intention to Recommend (AVE = 0.670; CR = 0.858)
I will recommend Sofia to my friends/family/colleagues 0.915 6.025 1.328 − 1.876 3.639
I will speak about my good impressions of Sofia to my friends/family/colleagues 0.917 6.121 1.250 − 2.095 4.822
I will be able to give helpful information about Sofia to my friends/family/colleagues 0.710 6.140 1.045 − 1.863 5.277
Intention towards Country’s Products (AVE = 0.717; CR = 0.910)
Once at home, I hope to be able to find Bulgarian products in local shops 0.861 3.459 1.069 − 0.394 − 0.425
Once at home, I would like to buy Bulgarian products 0.927 3.541 1.039 − 0.384 − 0.382
Once at home, I would be willing to search for Bulgarian products. 0.862 3.443 1.081 − 0.232 − 0.650
I will recommend to my friends/family to buy Bulgarian products. 0.774 3.793 1.020 − 0.646 − 0.033

Image construct), its convergent validity is still adequate. Therefore, the 4.3. Hypotheses testing
low AVE of Unique Image might be acceptable in this study since it does
not produce major reliability and discriminant validity problems. In fact, Table 4 presents the standardised and unstandardised coefficient
although the Fornell and Larcker’s (1981) criterion is not verified for estimates of the hypothesised direct effects between constructs. Table 5
some constructs (contrary to what was intended, the square root of the presents the standardised estimates of the indirect and total effects.
AVE values of Affective Image and Tourist Satisfaction are slightly lower These latter effects’ significance was estimated by the BC bootstrap
than the correlations between these latent variables and two other latent method (Cheung & Lau, 2008). All possible indirect effects are shown in
variables), it can be considered there is evidence of discriminant validity the table as a detail of the total indirect effects.
since the values are close from each other (see Table 3). Results show that the three components of the destination image
Whereas the skewness and kurtosis values for each item (see Table 2) have different effects on the model evaluative and behavioural vari­
are within the recommended values to assess the univariate normality ables. Only the perceptions of affective image people have about the
(lower than 3 and 7, respectively; Curran, West, & Finch, 1996), the destination directly influence both its satisfaction with the city (std.
multivariate kurtosis value of 17.934 confirms that the data are slightly coef. = 0.791; p < 0.01) and its intention to recommend (std. coef. =
departing from multivariate normality (although there is no consensual 0.242; p < 0.01), confirming the hypotheses H2 and H5 respectively.
threshold value for multivariate kurtosis, Bentler (2006) argues that it The cognitive image also directly influences the intention to recommend
should be lower than 5). For non-normal data, Bollen-Stine bootstrap the city, although this influence is the weakest (std. coef. = 0.138; p <
was used to provide the correct p-value for the chi-square statistic to 0.01); Hence, H4 is confirmed. The unique image only has an effect on
assess overall model fit (Bollen & Stine, 1992). The Bollen-Stine boot­ the intention towards the country’s products (std. coef. = 0.168; p <
strap p value is slightly lower than 0.01, which recommends the model 0.05), meaning that H10 is verified but H3 and H6 are not. Furthermore,
rejection. However, as Bollen-Sine p is very sensitive to sample size, the impact of the city’s affective image on recommendation intention is
researchers advise to use also other measures-of-fit as a criterion for strongly affected by the indirect effect through satisfaction (std. Indirect
model acceptance/rejection (Bollen & Stine, 1992). Therefore, the and total effects of 0.451 and 0.693, respectively), leading to tourist
overall model fit indices indicate an acceptable fit (X2 = 1081.657 df = satisfaction being a partial mediator of this relationship.
449, Bollen-Stine p = 0.0079; TLI = 0.887; CFI = 0.898; RMSEA = Regarding to the relationship between the evaluative and behav­
0.067, pClose<0.01; 90%CI=(0.062; 0.072)). It should be noted that ioural variables, the tourist satisfaction positively affects the intention of
SEM proves extremely robust with respect to slightly violations of its destination recommendation to others (std. coef. = 0.570; p < 0.01),
underlying assumptions (Reinartz, Haenlein, & Henseler, 2009). verifying H7, and indirectly the intention towards the country’s prod­
ucts although this latter effect is weak (std. Indirect eff. = 0.142; p <

7
C. Marques et al. Tourism Management 85 (2021) 104293

Table 3
Correlation values between constructs (a).
Cognitive Affective Unique Tourist Intention to Intention towards Country’s
Image Image Image Satisfaction Recommend Products

Cognitive Image 0.743

Affective Image 0.575 0.717
Unique Image 0.407 0.471 0.587
Tourist Satisfaction 0.479 0.707 0.318 0.722
Intention to Recommend 0.576 0.729 0.414 0.747 0.819
Intention towards Country’s 0.313 0.356 0.322 0.312 0.382 0.847
Products

(a) Diagonal values represent the square root of the construct AVE value.

Table 4
Coefficient estimates of the model structural relationships.
Hyp. Relationships Std estimate Unstd estimate Standard Error t Statistic

H1 Cognitive Image → Tourist Satisfaction 0.056 0.099 0.099 0.994

H2 Affective Image → Tourist Satisfaction 0.791* 0.713 0.064 11.156
H3 Unique Image → Tourist Satisfaction − 0.077*** − 0.142 0.103 − 1.380
H4 Cognitive Image → Intention to Recommend 0.138* 0.302 0.101 2.997
H5 Affective Image → Intention to Recommend 0.242* 0.269 0.083 3.258
H6 Unique Image → Intention to Recommend 0.062*** 0.141 0.103 1.365
H7 Tourist Satisfaction → Intention to Recommend 0.570* 0.702 0.084 8.324
H8 Cognitive Image → Intention towards Country’s Products 0.087 0.144 0.126 1.138
H9 Affective Image → Intention towards Country’s Products 0.059 0.049 0.105 0.471
H10 Unique Image → Intention towards Country’s Products 0.168** 0.288 0.131 2.195
H11 Tourist Satisfaction → Intention towards Country’s Products − 0.040 − 0.037 0.127 − 0.296
H12 Intention to Recommend → Intention towards Country’s Products 0.249** 0.189 0.109 1.730

*p < 0.01; **p < 0.05; ***p < 0.10.

0.05); Moreover, as there is no direct effect between tourist satisfaction However, cognitive image has no significant effect on tourist satisfac­
and intention towards the country’s products (H11 not verified), tion; unique image in turn, slightly negatively influences tourist satis­
intention to recommend the destination is thus a total mediator of this faction (− 0.077; p < 0.10), if a significance level of 10% be considered.
relationship. The intention towards country’s products is also directly These results agree partially with Chiu et al. (2016) who found that both
influenced by the intention to recommend the destination (std. coef. = affective and cognitive components influence satisfaction, although af­
0.249; p < 0.05; H12 verified), as well as by the unique image as seen fective image has a stronger effect on satisfaction than cognitive image.
above; In addition, it is indirectly influenced by the cognitive and af­ The reason why the cognitive image does not influence tourist satis­
fective images in a very weak way through all its indirect effects (std. faction in our study may be associated to the fact that the cognitive
total effects of 0.127 and 0.200, respectively). image items are related to the image held by tourists of accommodation,
Together, tourists’ destination image and satisfaction as predictors local food and gastronomy as infrastructural facilities for tourism which
explain a high proportion of variance in intention to recommend the are still widely underdeveloped. Our results however, are not consistent
destination (R2 = 0.782). Nevertheless, as far as the intention towards with those of others studies, perhaps due to the latter having different
country’s products is concerned, their antecedents only account for premises; for instance, Kozak and Rimmington (2000), Bigné et al.
18.5% of its variance. Tourist satisfaction, in turn is explained by 62.4% (2001), Chi and Qu (2008), Loureiro and González (2008), Ram­
of destination image and almost totally by affective image. seook-Munhurrun et al. (2015) and Aliman et al. (2016) found a positive
relationship between destination image and tourist satisfaction although
5. Discussion and Conclusion they have not decomposed destination image. Similarly, Chen and Tsai
(2007) results suggested that destination image has not a direct but an
This research aims to study destination image in the particular indirect effect on satisfaction through trip quality and perceived value.
context of emerging destinations. In this sense, it proposes and tests a In addition, Prayag et al. (2017) confirmed that both overall destination
conceptual model which aims to examine how the cognitive, affective, image and emotions arising from positive consumption experiences in­
and unique components of destination image influence the tourist fluence tourist satisfaction. Emotions have been defined as affective
satisfaction and its post-visit intentions to recommend the destination states charged with episodes of intense feelings related to a referent
and towards its products. Sofia, the capital of Bulgaria, was chosen as it (Prayag et al., 2017) and have been proven to influence satisfaction (e.g.
was ranked as the 3rd fastest growing city in Europe in terms of over­ Del Bosque & Martín, 2008). In the current study, affective images are
night visitors (Mastercard, 2017). Examining foreign tourist perceptions emotional associations made with the city of Sofia. Thus, a parallelism
of the image of the city upon visiting it, and post-visit behaviours that may be drawn between the effects of Sofia’s affective image on satis­
result from those perceptions, will elicit a better understanding what can faction and the emotions-satisfaction relationship of Prayag et al.
be improved in the marketing efforts of the local tourism institutions and (2017).
businesses. Regarding the determinants of intention to recommend, the results
Results from a sample of 314 international visitors to Sofia provide provide support for H4, H5, and H7. The affective destination image has
empirical support to most hypotheses formulated. The hypothesised the strongest influence on the intention of tourists to recommend Sofia
relationships between cognitive, affective, and unique images with (std. total effect of 0.693 (p < 0.01) through satisfaction, although the
tourism satisfaction (H1, H2, and H3) are not all verified. The results direct effect is 0.242 (p < 0.01), revealing that satisfaction is a partial
show that the affective component of destination image has a strong mediator of this relationship). WOM is also influenced by cognitive
impact on tourist satisfaction (0.791; p < 0.01; H2 is supported). image (direct and total effects of 0.138 (p < 0.01) and 0.170 (p < 0.05),

8
C. Marques et al. Tourism Management 85 (2021) 104293

Table 5
Standardised total and indirect effects.
Dependent Variables

Independent Variables Effects Tourist Satisfaction (TS) Intention to Intention towards

Recommend (IR) Country’s Products

Cognitive Image Total Effects 0.056 0.170** 0.127**

Total Indirect Effects – 0.032 0.040**
Indirect Effect via TS – 0.032 − 0.002
Indirect Effect via IR – – 0.034
Indirect Effect via TS and IR – – 0.008
Affective Image Total Effects 0.791* 0.693* 0.200*
Total Indirect Effects – 0.451* 0.141**
Indirect Effect via TS – 0.451* − 0.032
Indirect Effect via IR – – 0.060
Indirect Effect via TS and IR – – 0.112
Unique Image Total Effects − 0.077*** 0.018 0.176*
Total Indirect Effects – − 0.044 0.008
Indirect Effect via TS – − 0.044 0.003
Indirect Effect via IR – – 0.015
Indirect Effect via TS and IR − – − 0.011
Tourist Satisfaction Total Effects – 0.570* 0.102
Total Indirect Effects – – 0.142**
Indirect Effect via IR – – 0.142**
Intention to Recommend Total Effects – – 0.249**

*p < 0.01; **p < 0.05; ***p < 0.10.

respectively). These findings confirm Chiu et al. (2016) in which the effect of 0.142; p < 0.05). These results are consistent with De Nisco
authors found that the affective image has a direct influence on loyalty et al. (2016) who found that the higher the tourist satisfaction, the
whereas the cognitive image has only an indirect influence on loyalty higher the tourist willingness to recommend and buy Italian products.
through affective image and satisfaction. Likewise, Baloglu et al. (2014) Finally, the intention to recommend the city had the strongest direct
also found out that the affective component has a stronger effect on impact on the intentions towards Bulgarian products upon visiting Sofia
intention to recommend than the cognitive either for first time or repeat (0.249; p < 0.05).
visitors to Jamaica. Apart from the other two image components, per­ The theoretical contribution of the current work to the literature on
ceptions of the unique Sofia image also have a very small impact on tourism marketing is fourfold. All contributions are enunciated below,
WOM (0.062; p < 0.10), similarly to its impact on satisfaction (if using a together with the practical implications for the tourism of Sofia, an
significance level of 10%). The different degrees of strength of these emerging destination.
impacts on WOM also were obtained in Qu et al.’s (2011) study. In fact, Firstly, it is one of the few studies which analyses destination image,
even though every image component has an indirect influence on not only with its cognitive and affective components, but also adding
intention to recommend in the referred study, the cognitive image is the uniqueness. In addition, this research not only provides further insights
strongest component of image influencing WOM, followed by unique on the influence of unique image on intention to recommend the
image; The affective component is the one with the weakest influence, destination, but it is also the first study to test the influence of unique
contrary to tourists’ perceptions of Sofia in the current research. As for destination image on tourist satisfaction and post-visit intentions to­
tourist satisfaction, it has a strong impact on intention to spread positive wards products made in the visited country. Moreover, it also uses a new
word-of-mouth (0.570; p < 0.01), reiterating previous research (e.g. updated methodology to extract unique features of destinations based
Chen & Tsai, 2007; Chi & Qu, 2008; Kozak & Rimmington, 2000; Bigné on text-mining of UGC, in this case, blog posts about tourists’ experience
et al., 2001; Ozturk & Gogtas, 2016; Phillips et al., 2013; De Nisco et al., in Sofia. Nowadays, due to the amount of textual data available in Web,
2016; Prayag et al., 2017). text-mining of UGC has proved to be an effective method, as it not only
The findings regarding the determinants of intentions towards makes access to information easier and faster but also reflects vivid
recommendation and purchase of Bulgarian products support just two of visitor experiences.
the proposed relationships (H10 and H12). Firstly, although unique According to Qu et al. (2011), identifying the unique elements of a
destination image is the only image factor which directly influences destination should be the starting point of every positioning strategy as
post-visit intentions towards products made in Bulgaria (0.168; p < these elements are key differentiators. As this research shows, in the
0.05), the affective and cognitive images also have a positive total effect context of Sofia, unique image has a significant influence on tourist
on this construct through satisfaction and WOM (0.200 (p < 0.01) and post-visitation behaviour, namely on recommendation and purchasing
0.127 (p < 0.05), respectively). This finding is partially aligned with the intentions of Bulgarian products. As Sofia is an emerging tourist desti­
results of Elliot et al. (2011), perhaps because Sofia is not a well-known nation, policymakers need to promote its unique features, using imagery
destination. Using data from South Korean consumers attending a travel emphasizing interesting characteristics of the city, which makes it
show in Seoul in order to assess their images of Japan and the USA, the unique.
Elliot et al.’s (2011) study suggested that the affective country image Secondly, a conceptual model is put forth, adding several variables to
directly impacts product receptivity for both Japan and the USA, already tested frameworks, namely unique destination image and post-
although the effect was stronger for Japan. However, the cognitive visit intentions towards products in an emerging destination. Findings
image of these countries only influences product receptivity when show that affective image influences all the constructs, having the
mediated by product beliefs. This particular relationship is stronger for highest impact on tourist satisfaction and intention to recommend the
the USA than for Japan. Secondly, the current research results show that destination by traditional word-of-mouth, as other researchers have
the construct of intentions towards Bulgarian products is not directly shown for mature destinations. Additionally, a path of relationships is
influenced by tourist satisfaction, but through positive WOM (indirect derived from the parameter estimates: affective image – tourist

9
C. Marques et al. Tourism Management 85 (2021) 104293

satisfaction – intention to recommend the destination – intention to­ unique characteristics. In so doing, it follows that the promotion of Sofia
wards country’s products. These suggest that the way tourists feel during as an emerging destination is required to emphasize what are its unique
their stay is the most influential factor in their travel experience. natural and cultural attractions, as well as ensuring that more up to date
Therefore, as tourism is based on experiences, tourism businesses should infrastructure for visitor accommodation exists.
be focussed on creating offerings which are able to invoke a positive
affective evaluation in consumers, as this will lead to a better overall 6. Limitations and recommendations for future research
image, increased satisfaction levels, and positive word-of-mouth offline
and online. Electronic word-of-mouth could be especially beneficial This study has some limitations that should be listed. The first lim­
since it does not only provide positive feedback about the image of Sofia, itation is related to the sample composition. Some of the age groups,
but also of the services being offered there, such as tours, hotels, res­ namely 65 or older and 55–64, are under-represented in the study.
taurants, etc. Regarding the cognitive component of destination image Future research should use a sample including a more represented re­
of Sofia, it has a significant impact on all the constructs, except on tourist spondents from these age groups. Another limitation is that the ques­
satisfaction. It is the least important destination image component to tionnaire was only in English. During the data collection, there were
influence post-visit intentions towards Bulgarian products but is more several occasions when tourists visiting Sofia could not fill out the sur­
influential than unique image in recommendation intention. Cognitive vey because they were unable to understand. The questionnaire needs to
image had little impact on the behavioural outcomes, perhaps because it be translated into different languages, based on statistical data about the
referred to attributes related to local cuisine, accommodation and local country-of-origin of the tourists. A third limitation is related to the low
products. As alluded to earlier, these results may be attributable to Sofia AVE value of the Unique Image factor. Although the AVE value could be
being an emerging destination whose image has not yet been improved by dropping items with large measurement error variance, it
consolidated. could result in an item set of higher AVE value which could be less
Thirdly, this is one of the few studies which empirically tests the content or face valid than before items were dropped. Furthermore, as
influence of the different components of destination image on tourists’ (1) Fornell and Larcker (1981) argue that the convergent validity is not
post-visit intentions towards products made in the visited country (De questionable when the factor is strongly reliable, (2) some researchers
Nisco et al., 2015; Elliot et al., 2011). The study links two streams of do not consider AVE as a measure of convergent validity, but rather
research, COI and TDI, integrating tourism with country’s products, thus prefer reliability and (3) the unique image factor and its items demon­
exploring the crossover relationships between them. Understanding strate strong reliability, it was decided to keep the unique image factor
these relationships are important since there are practical implications composed by eight items and a low AVE value. However, the resulting
for policymakers dealing with tourism. Whilst post-visit propensity to unique image factor is provisional, and more measurement work is
comment favourably on a destination will have a direct impact on tourist needed. The fourth limitation is that the three components of destination
travel revenue and subsequent consumption in the destination, it is image were measured in the final model with very few items. As Sofia is
conceivable that a willingness to buy destination country’s products an emerging destination, tourists have different views on destination
implies continuous consumption in the home country of products image. Some items consequently showed low intra-correlations and
experienced in the destination, with obvious favourable repercussions were thus removed from the overall measurement model. The items that
on exports and the balance of trade of the destination country. The remain in the model are those in which diverse sources of information
current study shows that all destination image components significantly converge (Kim & Chen, 2016) allowing tourists to form a common image
influence post-visit intentions towards Bulgarian products, but affective of Sofia.
image has the strongest impact. Furthermore, unique image has a direct Future research should test the proposed conceptual model in
positive influence on the intentions towards products after visiting the different emerging capital cities in Europe and compare it with the
country, regardless of any other variable. As the unique image items current findings. Other constructs could be incorporated in the model to
resulting from the CFA focus essentially on a multicultural city with its further examine the relationships between destination image and loyalty
own peculiarities like the open book market, squat shops, etc., it is not outcomes such as trip quality and perceived value. The method of
surprising to find a relationship to Bulgarian products. This should be studying unique destination image by text-mining of UGC (whether it is
considered by tourism professionals and officials in the way they present blog posts, website reviews, social media posts, etc.) should be repli­
the local products to international tourists. Finally, this is the first study cated for other cities. For instance, studying neighbouring capital cities
to test the relationship between the intention to spread positive WOM (Bucharest, Athens, Skopje, Belgrade) could provide insights on how
and post-visit intentions to recommend and buy Bulgarian products. they differ and what makes each unique.
Results show that only the traditional WOM has a significant effect on
the intention to recommend and buy Bulgarian products once the Contributions
tourists are back home.
Finally, this study highlights the idea that in an emerging destination Catarina Marques: Introduction and main definition of the Study
all the image components contribute in different ways to image and Aim, Methodology and Sampling Definition, Statistical Analyses, Dis­
branding destination (Qu et al., 2011). The affective image of an cussion and Conclusion, Rui Vinhas da Silva: Main definition of the
emerging destination must be noted in creating the destination identity Study Aim, Literature Review, Conclusion, Stanislava Antova: Intro­
and used for its positioning strategy as suggested by Baloglu and Brin­ duction and definition of the Study Aim, Literature Review, Collection of
berg (1997). The cognitive image component should gradually be Data, Text-Mining, Discussion.
created in the tourist’s mind, since it also has a strong influence on
post-visit behaviours as has been evidenced in research on mature des­ Declarations of competing interest
tinations. The importance of unique image, on the other hand, is related
to its usefulness in positioning the destination brand and differentiating None.
it from other potential destinations (Echtner & Ritchie, 1993; Qu et al.,
2011). In particular, regarding to its effects on intentions towards the Acknowledgements
country’s products, the higher the unique image tourists have, the
higher will be their intention to recommend and purchase the country’s None.
products and the higher the impact on country exports. In establishing
an emerging destination in the minds of potential visitors, policymakers
are required to put forth and institutionally communicate what are its

10
C. Marques et al. Tourism Management 85 (2021) 104293

Appendix A. Supplementary data tourism research association: Advancing tourism research globally (Vol. 16). Retrieved
from: https://scholarworks.umass.edu/ttra/2016/Academic_Papers_Visual/16.
De Nisco, A., Mainolfi, G., Marino, V., & Napolitano, M. R. (2015). Tourism satisfaction
Supplementary data related to this article can be found at https:// effect on general country image, destination image, and post-visit intentions. Journal
doi.org/10.1016/j.tourman.2021.104293. of Vacation Marketing, 21(4), 305–317.
Del Bosque, I. R., & Martín, H. S. (2008). Tourist satisfaction: A cognitive-affective
model. Annals of Tourism Research, 35, 551–573.
References Echtner, C., & Ritchie, B. (1991). The meaning and measurement of destination image.
Journal of Tourism Studies, 2, 2–12.
Aliman, N. K., Hashim, S. M., Harudin, S., & Wahid, S. D. M. (2016). Tourists’ satisfaction Echtner, C., & Ritchie, B. (1993). The measurement of destination image: An empirical
with a destination: An investigation on visitors to Langkawi Island. International assessment. Journal of Travel Research, 31(4), 3–13.
Journal of Marketing Studies, 8(3), 173–188. Elliot, S., Papadopoulos, N., & Kim, S. S. (2011). An integrative model of place image:
Anderson, J., & Gerbing, D. (1988). Structural equation modeling in practice: A review Exploring the relationships between destination, product, and country images.
and recommended two-step approach. Psychological Bulletin, 103(3), 411–423. Journal of Travel Research, 50(5), 520–534.
Arslanova, L., Agapito, D., & Pinto, P. (2017). Tourism destination image of Russia in the Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with
perspective of the Portuguese market. Journal of Spatial and Organizational Dynamics, unobservable variables and measurement error. Journal of Marketing Research, 18(1),
5(2), 146–158. 39–50.
Aureli, S. (2017). A comparison of content analysis usage and text mining in CSR Gunn, C. A. (1972). Vacationscape: Designing tourist regions. Washington: Taylor &
corporate disclosure. International Journal of Digital Accounting Research, 17, 1–32. Francis.
Bagozzi, R., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of Hair, J. F. J., Black, W. C., Babin, B. J., & Anderson, R. E. (2014). Multivariate data
the Academy of Marketing Science, 16, 74–94. analysis (7th ed.). Essex, UK: Pearson New International Edition.
Baloglu, S. (1999). A path analytical model of visitation intention involving information Hallberg, A. (2005). Post-travel consumption country-of-origin effects of international travel
sources, socio-psychological motivations and destination images. Journal of Travel & experiences. Doctoral PhD dissertation. Sweden: University of Gothenburg.
Tourism Marketing, 8(3), 81–90. Hosany, S., Ekinci, Y., & Uysal, M. (2007). Destination image and destination personality.
Baloglu, S., & Brinberg, D. (1997). Affective images of tourism destinations. Journal of International Journal of Culture, Tourism and Hospitality Research, 1(1), 62–81.
Travel Research, 35(4), 11–15. Hunt, J. (1975). Image as a factor in tourism development. Journal of Travel Research, 1
Baloglu, S., Henthorne, T. L., & Sahin, S. (2014). Destination image and brand (3), 1–7.
personality of Jamaica: A model of tourist behaviour. Journal of Travel & Tourism Hunt, C., & Stronza, A. (2014). Stage-based tourism models and resident attitudes
Marketing, 31(8), 1057–1070. towards tourism in an emerging destination in the developing world. Journal of
Baloglu, S., & McCleary, K. (1999). A model of destination image formation. Annals of Sustainable Tourism, 22(2), 279–298.
Tourism Research, 26(4), 868–897. Kim, H., & Chen, J. S. (2016). Destination image formation process: A holistic model.
Basaran, U. (2016). Examining the relationship of cognitive, affective, and conative Journal of Vacation Marketing, 22(2), 154–166.
destination image: A research on safranbolu, Turkey. International Business Research, Kim, E. E. K., Mattila, A. S., & Baloglu, S. (2011). Effects of gender and expertise on
9(5), 164–179. consumers’ motivation to read online hotel reviews. Cornell Hospitality Quarterly, 52,
Beerli, A., & Martín, J. D. (2004). Factors influencing destination image. Annals of 399–406.
Tourism Research, 31(3), 657–681. Kim, S. S., Mckercher, B., & Lee, H. (2009). Tracking tourism destination image
Bentler, P. M. (2006). EQS 6 structural equations program manual. Encino: CA: perception. Annals of Tourism Research, 36(4), 715–718.
Multivariate Software, Inc. Konecnik, M., & Gartner, W. C. (2007). Customer-based brand equity for a destination.
Bigné, J. E., Andreu, L., & Gnoth, J. (2005). The theme part experience: An analysis of Annals of Tourism Research, 34(2), 400–421.
pleasure, arousal and satisfaction. Tourism Management, 26, 833–844. Kotler, P., & Gertner, D. (2004). Country as brand, product and beyond: A place
Bigné, J. E., Sanchez, M. I., & Sanchez, J. (2001). Tourism image, evaluation variables marketing and brand management perspective. In N. Morgan, A. Pritchard, &
and after purchase behaviour: Inter-relationship. Tourism Management, 22(6), R. Pride (Eds.), Destination branding: Creating the unique destination proposition (2nd
607–616. ed.). Oxford, UK: Elsevier, Ltd.
Bilkey, W. J., & Nes, E. (1982). Country-of-origin effects on product evaluations. Journal Kotler, P., Haider, D. H., & Rein, I. (1993). Marketing places. New York: The Free Press.
of International Business, 13(1), 89–99. Kozak, M., & Rimmington, M. (2000). Tourist satisfaction with Mallorca, Spain, as an off-
Bollen, K. A., & Stine, R. (1990). Direct and indirect effects: Classical and bootstrap season holiday destination. Journal of Travel Research, 38(3), 260–269.
estimates of variability. Sociological Methodology, 20, 115–140. Kozinets, R. V. (2002). The field behind the screen: Using netnography for marketing
Bollen, K. A., & Stine, R. (1992). Bootstrapping goodness-of-fit measures in structural research in online communities. Journal of Marketing Research, 39(1), 61–72.
equation models. Sociological Methods & Research, 21(2), 205–229. Leach, M. P., Liu, A. H., & Winsor, R. D. (2008). The impact of attitudes, word-of-mouth,
Butler, R. W. (1980). The concept of a tourism areas cycle of evaluation: Implications for and value congruence on conference participation: A comparison of attending and
management of resources. Canadian Geographer, 24(1), 5–12. non-attending organizational members. Journal of Hospitality Marketing &
Chadee, D., & Mattsson, J. (1996). An empirical assessment of customer satisfaction in Management, 16(3), 246–269.
tourism. The Services Industrial Journal, 16(3), 305–320. Lee, R., & Lockshin, L. (2012). Reverse country-of-origin effects of product perceptions
Chattalas, M., Kramer, T., & Takada, H. (2008). The impact of national stereotypes on the on destination image. Journal of Travel Research, 51(4), 502–511.
country of origin effect A conceptual framework. International Marketing Review, 25 Lopes, S. D. F. (2011). Destination image: Origins, developments and implications.
(1), 54–74. Revista de Turismo Património Cultural, 9(2), 305–315.
Chen, Y.-F., & Law, R. (2016). A Review of research on electronic word-of-mouth in Loureiro, S. M. C., & González, F. J. M. (2008). The importance of quality, satisfaction,
hospitality and tourism management. International Journal of Hospitality & Tourism trust, and image in relation to rural tourist loyalty. Journal of Travel & Tourism
Administration, 17(4), 347–372. Marketing, 25(2), 117–136.
Chen, C., & Phou, S. (2013). A closer look at destination image: Image, personality, MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the
relationship and loyalty. Tourism Management, 36, 269–278. indirect effect: Distribution of the product and resampling methods. Multivariate
Chen, C. G., & Tsai, D. (2007). How destination image and evaluative factors affect Behavioral Research, 39, 99–128.
behavioural intentions. Tourism Management, 28, 1115–1122. Marchiori, E., & Onder, I. (2017). A comparison of pre-visit beliefs and projected visual
Cheung, G. W., & Lau, R. S. (2008). Testing mediation and suppression effects of latent images of destinations. Tourism Management Perspectives, 21, 42–53.
variables: Bootstrapping with structural equation models. Organizational Research Martínez Caro, L., & Martínez García, J. A. (2007). Cognitive-affective model of
Methods, 11(2), 296–325. consumer satisfaction. An exploratory study within the framework of a sporting
Chevalier, J., & Mayzlin, D. (2006). The effect of word of mouth on sales: Online book event. Journal of Business Research, 60(2), 108–114.
reviews. Journal of Marketing Research, 43, 345–354. Mastercard. (2017). Mastercard destination cities Index. Retrieved from https://newsroom.
Chiadmi, N. E., Ladhari, R., & Souiden, N. (2017). Destination personality and mastercard.com/wp-content/uploads/2017/09/Mastercard-Destina-tion-Cities-Ind
destination image. Journal of Hospitality and Tourism Management, 32, 54–70. ex-Report.pdf. (Accessed 10 September 2018).
Chi, C. G.-Q., & Qu, H. (2008). Examining the structural relationships of destination Murphy, L., Mascardo, G., & Benckendorff, P. (2007). Exploring word-of-mouth
image, tourist satisfaction and destination loyalty: An integrated approach. Tourism influences on travel decisions: Friend sand relatives vs. Other travellers. International
Management, 29, 624–636. Journal of Consumer Studies, 21, 517–527.
Chiu, W., Zeng, S., & Cheng, P. S. (2016). The influence of destination image and tourist Oliver, R. (1993). Cognitive, affective and attribute bases of the satisfaction response.
satisfaction on tourist loyalty: A case study of Chinese tourists in korea. International Journal of Consumer Research, 20, 418–430.
Journal of Culture, Tourism and Hospitality Research, 10(2), 223–234. Ozturk, U. A., & Gogtas, H. (2016). Destination attributes, satisfaction, and the visitor’s
Confente, I. (2015). Twenty-five years of word-of-mouth studies: A critical review of intent to revisit and recommend. Tourism Geographies, 18(2), 194–212.
tourism research. International Journal of Tourism Research, 17(6), 613–624. Papadopoulos, N., & Heslop, L. A. (1986). Travel as a correlate of product and country
Crompton, J. L., Fakeye, P. C., & Lue, C.-C. (1992). Positioning: The example of the lower images. In T. E. Muller, & Marketing (Eds.), Canada: Marketing division, administrative
rio grande valley in the winter long stay destination market. Journal of Travel sciences association of Canada (pp. 191–200).
Research, 31(2), 20–26. Park, S., & Allen, J. P. (2013). Responding to online reviews: Problem solving and
Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to engagement in hotels. Cornell Hospitality Quarterly, 54(1), 64–73.
nonnormality and specification error in confirmatory factor analysis. Psychological Patterson, I. (2007). Information sources used by older adults for decision making about
Methods, 1, 16–29. tourist and travel destinations. International Journal of Consumer Studies, 31,
De Nisco, A., Elliot, S., & Papadopoulos, N. (2016). Country image, tourist satisfaction, 528–533.
and future visit and product purchase intentions: An integrative model. In Travel and

11
C. Marques et al. Tourism Management 85 (2021) 104293

Peter, J. P., & Olson, J. C. (1999). Consumer behavior and marketing strategy. Boston: Irwin
McGraw-Hill.
Phillips, W. J., Wolfe, K., Hodur, N., & Leistritz, F. L. (2013). Tourist word of mouth and CATARINA MARQUES holds a PhD in Quantitative Methods
revisit intentions to rural tourism destinations: A case of north Dakota, USA. with specialization in Statistics and Data Analysis. She is As­
International Journal of Tourism Research, 15, 93–104. sistant Professor of Statistics and Data Analysis at the Depart­
Prayag, G., Hosany, S., Muskat, B., & Del Chiappa, G. (2017). Understanding the ment of Quantitative Methods for Management and Economics,
relationships between tourists’ emotional experiences, perceived overall image, ISCTE Business School, Portugal. Her current research interests
satisfaction, and intention to recommend. Journal of Travel Research, 56(1), 41–54. are in statistical modeling and data analysis, especially in
Qu, H., Kim, L. H., & Im, H. H. (2011). A model of destination branding: Integrating the structural equation modeling and clustering methods. She has
concepts of the branding and destination image. Tourism Management, 32, 465–476. taught Data Analysis on several Master and PhD programs,
Racherla, P., Connolly, D. J., & Christodoulidou, N. (2013). What determines consumers’ namely in a Tourism Management PhD program. She has
ratings of service providers? An exploratory study of online traveler reviews. Journal published in Journal of Sustainable Tourism, Annals of
of Hospitality Marketing & Management, 22(2), 135–161. Tourism Research, Journal of Destination Marketing & Man­
Ramseook-Munhurrun, P., Seebaluck, V. N., & Naidoo, P. (2015). Examining the agement, among others.
structural relationships of destination image, perceived value, tourist satisfaction
and loyalty: Case of maritius. Procedia Social and Behavioural Sciences, 175, 252–259.
Reinartz, W., Haenlein, M., & Henseler, J. (2009). An empirical comparison of the
efficacy of covariance-based and variance-based SEM. International Journal of
Research in Marketing, 26(4), 332–344.
Russell, J. A., & Pratt, G. A. (1980). A description of the affective quality attributed to RUI VINHAS DA SILVA holds a PhD and Post-Doctorate from
environments. Journal of Personality and Social Psychology, 38, 311–322. Manchester Business School, UK. He is an Associate Professor
Sigala, M., Christou, E., & Gretzel, U. (2012). Social media in travel, tourism and hospitality; at ISCTE Business School, Portugal. Currently he is Executive
theory, practice and cases. London: Ashgate. Board Member at INDEG-ISCTE. He was Chairman of the Board
Sotiriadou, P., Brouwers, J., & Le, T. A. (2014). Choosing a qualitative data analysis tool: at COMPETE 2020 (Dec. 2014–May 2016). Before joining
A comparison of NVivo and leximancer. Annals of Leisure Research, 17(2), 218–234. ISCTE in April 2010, he was an Associate Professor (Senior
Stoilova, Z. (2016). Дискрет ният т урист ически буМ в ?офия [the discreet tourism boom in Lecturer) (2008) at MBS (Assistant Professor since 1998). He
Sofia]” newspaper capital. https://www.capital.bg/politika_i_ikonomika/obshtestvo/ has been taught on MBA and PhD programmes. Vinhas da Silva
2016/11/20/2866308_diskretniiat_turisticheski_bum_v_sofiia/. (Accessed 1 has interests in national competitiveness, country branding,
September 2018). reputation and country-of-origin effects. He has published
Stringam, B. B., & Gerdes, J., Jr. (2010). An analysis of word-of-mouse ratings and guest several books and articles.
comments of online hotel distribution sites. Journal of Hospitality Marketing &
Management, 19(7), 773–796.
Stylos, N., Vassiliadis, C. A., Bellou, V., & Andronikidis, A. (2016). Destination images,
holistic images and personal normative beliefs: Predictors of intention to revisit a
destination. Tourism Management, 53, 40–60.
Talib, R., Hanif, M. K., Ayesha, S., & Fatima, F. (2016). Text mining techniques, STANISLAVA ANTOVA holds a MSc in Marketing from ISCTE
applications and issues. Journal of Advanced Computer Science and Applications, 7(11), Business School. She is currently working as advertisers ac­
414–418. count manager in the martech industry. Her research interests
Tasci, A. D. A., Gartner, W. C., & Cavusgil, S. T. (2007). Conceptualization and are in destination image, destination branding, customer
operationalization of destination image. Journal of Hospitality & Tourism Research, 31 satisfaction and post-visit intentions.
(2), 194–223.
Um, S., & Crompton, J. L. (1990). Attitude determinants in tourism destination process.
Annals of Tourism Research, 17(3), 432–448.
Westbrook, R. A. (1980). Interpersonal affective influences on consumer satisfaction with
products. Journal of Consumer Research, 7, 49–54.
Wong, L. (2008). Data analysis in qualitative research: A brief guide to using NVivo.
Malaysian Family Physician, 3(1), 14–20.
World Population Review. (2017). Bulgaria population. Retrieved from http://world
populationreview.com/countries/bulgaria. (Accessed 9 September 2018).
Yoon, Y., & Uysal, M. (2005). An examination of the effects of motivation and satisfaction
on destination loyalty: A structural model. Tourism Management, 26, 45–56.

12